Visual field loss is a common corollary of stroke (Pollock et al,The Cochrane Library:John Wiley & Sons; 2011,p p. 1-83) that is prevalent in the VA population. Posterior and middle cerebral artery infarcts often injure visual cortical networks resulting in partial or complete homonymous hemianopia or quadrantanopia. The most common clinically significant visual cortical injury involves the primary visual cortex and adjacent regions (area V1+). Area V1 is the chief relay of visual input to higher (extrastriate) cortical areas, and V1 lesions result in a dense contralateral scotoma within which visual perception is severely impaired. The resulting visual deficit is long thought to be resistant to rehabilitation, i.e. to be essentially irreversible, and impairs quality of life significantly. Visual field loss impairs many activities of daily living, including the ability to ambulate safely, drive, read, cook, supervise minors, and others, markedly reducing the independence of the patient (Veterans Health Initiative 2002: Visual Impairment and Blindness). Loss of visual motion perception is particularly problematic for navigating the environment and for avoiding collision with moving objects. Here we focus on studying rehabilitation of visual motion perception. Several studies demonstrated that visual cortex can be driven from the interior of the scotoma, i.e. from areas of the visual field where there is no visual perception. One such area that can be visually modulated in the absence of V1 input is hV5/MT+, an area important for visual motion perception. Although this modulation is weak and does not generally confer useful vision (?blindsight?), it raises the hope that appropriately designed new rehabilitative strategies may be able to strengthen these partially functioning pathways promoting recovery. Recent work (Huxlin et al. J Neuroscience 29(13):3981-91, 2009) has shown that intensive training in a visual motion discrimination task can improve visual motion perception in some subjects with cortical V1 lesions (see fig 5-6). However, this work is still at the preliminary stage. Specifically, not all subjects appear to benefit (section C2), it is not clear which parts of the visual field are amenable to visual rehabilitation, and the mechanism of recovery remains unclear. Our goal here is twofold: 1) To pilot functional magnetic resonance imaging (fMRI) population receptive field mapping methods for identifying regions of the visual field that are more amenable to visual rehabilitation after lesions of the primary visual cortex (area V1+). 2) To probe the mechanisms underlying the recently reported rehabilitation-induced improvement in visual performance that occurs in some subjects (preliminary data fig. 5-7). We will address the following specific aims: Aim 1: We will show that functional magnetic resonance imaging (fMRI) population receptive field mapping methods can identify regions of the visual field that are more amenable to visual motion perception rehabilitation after lesions of the primary visual cortex (area V1+). Hypothesis 1: Regions of the scotoma that elicit visually driven fMRI activity in both spared area V1 and hV5/MT+ (see fig.1) will be easier to rehabilitate, and will reach higher behavioral thresholds, compared to regions that activate only area hV5/MT+. Aim 2: Study how the strength of visual modulation changes in early visual areas, in hV5/MT+, and in fronto- parietal networks subserving attention & higher-order visual processing, following rehabilitative training. Determine whether observed changes are correlated with recovery. Hypothesis 2: Training will increase the signal to noise ratio of the response to visual motion stimuli in area hV5/MT+, while responses in spared V1+ and early extrastriate areas will remain relatively unchanged. Increase in hV5/MT+ response strength will correlate best with recovery. The change in hV5/MT+ response strength will likely be mediated via top down pathways, potentially associated with attentional networks (see [43] [44] and preliminary data fig. 5-7). Our long-term goal is to use information obtained from fMRI methods to implement a principled strategy for developing rehabilitative treatments following visual cortex injury.